The Holy Grail of Segmentation

"The right message, to the right person, at the right time" - we've all heard it.

?

It's time to talk about segmentation (again), ladies & gents.

?

I've spoken a huge amount about this in the past (read here) so I won't beat around the bush today but focus more on my core beliefs as well as how segmentation will change in the coming years.

?

First, let me say that I've been probably the worst offender for overcomplicating segmentation in the past and spread a lot of poor information.

?

It's always been done with the best intentions to improve open rates, click rates, yada yada yada.

?

But these days my views are more simplified (but also becoming more irrelevant due to technological changes happening).

?

Let me explain.

?

First of all, the goal with segmentation isn't to improve click rates or open rates - it's to maximise revenue.

?

Open rates and click rates improving are just bi-products of better personalisation.

?

What you're really aiming for when you segment your audience is to drive the most profitable results possible.

?

Note that profitable above is bolded.

?

That's because the strategies email marketers create are often at odds with what's best for the brand (and also themselves).

?

The ideal segmentation strategy would pull a mixture of zero-party data and transactional data from your store's history to maximise revenue generation.

?

Marketers have tried their best to guess their way to implementing this strategy effectively, but there's a major bottleneck towards achieving it.

?

Every additional email you send requires new copywriting, strategy and design, as well as Q/A.

?

In simple words: it's labor cost that needs to be paid out for by the brand and measured against its true impact on profits.

?

So whilst the theory is great, the practicality and realism of implementing this falls short for the overwhelming majority of brands.

?

Because even if you generate 20% additional revenue by increasing labor costs by 300% per campaign, that's usually going to result in a significantly less profitable email program for most brands until they're operating at serious, serious scale.

?

You're also introducing more convoluted workflows into the mix that zap creative energy.

?

Is it worth it? For 99% of brands, no, it's not.

?

That's why I've said time and time again, for the overwhelming majority of the time, just go broad with your sends without screwing up your deliverability.

?

But the exciting news is that this strategy is going to be possible in the near future.

?

Technology is going to fuse data and creative together that makes this cumbersome workflow redundant.

?

It's difficult to say when this day will come, but it's evident that all major ESPs (including Klaviyo & Yotpo) are making moves down this route.

?

AI-based segment builders are already on the block, and it's inevitable that tools like ChatGPT & Midjourney will plug in to the software to generate the creatives.

?

This will be great for brands, but potentially?really bad for agencies.

?

Software will cut out the middlemen and brands will be more profitable.

?

At least that's their hope.

?

Personally, I envision a future where agencies will just be leaner, more specialised in prompting (and data analysis) and leveraging the tech intelligently.

?

That day is coming sooner rather than later and everybody should prepare for it.

Speaking of segmentation, it's always a good time to revisit this article I wrote a while back now.

?

I still stand by this piece and think most people abuse 'retention' channels in a grossly inefficient way.

?

Cannibalising email revenue with SMS and now even being too trigger happy with postcards can quickly rack up costs in a way that doesn't make sense to the brand.

?

Follow this messaging hierarchy and you'll always be at your most profitable.